One-step synthesis of 5-(4-fluorobenzyl)-2-furyl methyl ketone: A key intermediate of HIV-integrase inhibitor S-1360

Article abstract of DOI:10.1021/op700117q

Practical one-step synthesis of 5-(4-fluorobenzyl)-2-furyl methyl ketone was accomplished by Friedel-Crafts benzylation of 2-furyl methyl ketone with 4-fluorobenzyl chloride hi the presence of ZnCl₂. Two reaction conditions and their work-up pr

Full text of DOI:10.1021/op700117q

Organic Process Research & Development 2007, 11, 1059–1061
Communications to the Editor
One-Step Synthesis of 5-(4-Fluorobenzyl)-2-furyl Methyl Ketone: A Key Intermediate
of HIV-Integrase Inhibitor S-1360
Kenji Izumi,‡ Mikio Kabaki,† Masaaki Uenaka,† and Sumio Shimizu*,†
Chemical DeVelopment Department, CMC DeVelopment Laboratories, Shionogi & Co., Ltd., 1-3, Kuise Terajima 2-Chome,
Amagasaki, Hyogo 660-0813, Japan
Scheme 1. Medicinal chemistry route of S-1360
Abstract:
Practical one-step synthesis of 5-(4-fluorobenzyl)-2-furyl methyl
ketone was accomplished by Friedel–Crafts benzylation of 2-furyl
methyl ketone with 4-fluorobenzyl chloride in the presence of
ZnCl2. Two reaction conditions and their work-up procedures are
reported. The first utilizes an anhydrous condition in dichlo-
romethane, providing a convenient procedure for the isolation of
the product. The second involves an aqueous condition, which
offers a large-scale ecological and safe manufacturing process.
Introduction
5-(4-Fluorobenzyl)-2-furyl methyl ketone (5) is a key
intermediate in the synthesis of HIV-integrase inhibitor S-1360,
which was found in the Discovery Research Laboratory of
Shionogi & Co., Ltd.1 The HIV integrase has been considered
an attractive therapeutic target, since HIV cannot replicate
without integration in a host chromosome.2 To develop S-1360,
we needed a low-cost and reliable synthetic method by which
S-1360 could be prepared on a large scale.3 Here we report
two procedures for the one-step synthesis of 5 via Friedel–Crafts
benzylation.
2-furoic acid (3), which can be purified by crystallization.
The acid 3 was converted into the corresponding Wein-
reb’s amide5 4 and allowed to react with methylmagnesium
chloride to give ketone 5.
Results and Discussion
The medicinal chemistry route of S-1360 is shown in
Scheme 1. A few kilograms of ketone 5 were synthesized
by this route for nonclinical use. 2-Furoic acid (1) was
lithiated with 2 molar equiv of LDA in the presence of
HMPA, and the obtained dilithio derivative reacted with
4-fluorobenzaldehyde to give an alcohol 2. This alcohol
was reduced by TMSCl/NaI4 to afford 5-(4-fluorobenzyl)-
This chemistry presented the following problems for
industrial manufacturing of 5. (1) Toxic HMPA is needed
to dissolve the generated dilithio derivatives. (2) Iodine
has to be recovered in order to protect the environment.
(3) LDA, MeNHOMe · HCl, HOBt and MeMgCl are
expensive. To eliminate these scale-up issues, we inves-
tigated other synthetic methods. Friedel–Crafts-type reac-
tions of electron-rich furans6 and electron-deficient furans7
are well known,8 but there are few examples of furyl
ketones. The exception is the Friedel–Crafts alkylation
of 2-furyl methyl ketone (7),9 which led us to examine
the Friedel–Crafts benzylation of 7.
* To whom correspondence should be addressed. E-mail: sumio.shimizu@
shionogi.co.jp.
† Chemical Development Department, CMC Development Laboratories.
‡ Current address: Business Process Improvement Unit, Manufacturing
Division, Shionogi & Co., Ltd., 1-3, Kuise Terajima 2-Chome, Amagasaki,
Hyogo 660-0813, Japan.
(1) Fujishita, T.; Yoshinaga, T.; Sato, A. PCT Int. Appl. WO-0039086
A, 2000.
(2) (a) De Clercq, E. Expert Opin. Emerging Drugs 2005, 10, 241. (b)
De Clercq, E. Med. Res. ReV 2002, 22, 531. (c) Gulick, R. M.;
Staszewski, S. AIDS 2002, 16, S135.
Method A. We developed an excellent synthetic method
of 5, which involves Friedel–Crafts reaction of 7 with
(3) Shimizu, S.; Endo, T.; Izumi, K.; Mikamiyama, H. Org. Process Res.
DeV. 2007, 11, 1055.
(5) (a) NahmS.; Weinreb, S. M. Tetrahedron Lett. 1981, 22, 3815. (b)
Luca, L. D.; Giacomelli, G.; Taddei, M. J. Org. Chem. 2001, 66, 2534.
(c) Fehrentz, J.-A.; Castro, B. Synthesis 1983, 676. (d) Sibi, M.P. Org.
Prep. Proced. Int. 1993, 25, 15.
(4) (a) Sakai, T.; Miyata, K.; Utaka, M.; Takeda, A. Tetrahedron Lett.
1987, 28, 3817. (b) Sakai, T.; Miyata, K.; Tsuboi, S.; Takeda, A.;
Utaka, M.; Tori, S. Bull. Chem. Soc. Jpn. 1989, 62, 3537.
10.1021/op700117q CCC: $37.00
Published on Web 10/10/2007
2007 American Chemical Society
Vol. 11, No. 6, 2007 / Organic Process Research & Development
•
1059

Scheme 2. Method A
Scheme 3. Method B
layer system, which contains a layer of organic compounds and
aqueous ZnCl2 solution. Namely, the mixture of 7, 2 molar
equiv of 8, and 1.05 molar equiv of ZnCl2 in 2 vol of water
was stirred at 85 °C for 6 h. The Friedel–Crafts reaction gave
70%10 of 5 accompanyied by 5%10 of ketone 9 and 5%10 of
di-4-fluorobenzyl ether (10). In order to remove the isomer 9
and the ether 10, the isolation procedure of 5 needed distillation
under reduced pressure and then crystallization under -10 °C
from mixed solvents of isopropyl alcohol and n-hexane. The
desired ketone 5 was isolated in 43% yield from 7. The
calorimeter study revealed the total energy given off during this
reaction with H2O to be 23.7 kJ/mol, which corresponded to
an adiabatic temperature rise of 14 °C. This condition was
adopted for large-scale synthesis and we were able to obtain
ca. 500 kg/lot of ketone 5 successfully.
p-fluorobenzylchloride (8) in the presence of anhydrous
ZnCl2 in dichloromethane to afford 5 in 53%10 yield
accompanied by 1%10 of 4-(4-fluorobenzyl)-2-furyl methyl
ketone (9) (Scheme 2). This one-step synthesis presents
the following advantages. The starting materials 7 and 8
are inexpensive, and the isolation procedure of 5 is
convenient. Namely, a mixture of 7, 2 molar equiv of 8,
1.5 molar equiv of anhydrous ZnCl2 powder and dichlo-
romethane was stirred at 45 °C. As the reaction proceeded,
a complex with 5 and ZnCl2 precipitated in dichlo-
romethane. The precipitates were collected by filtration
to remove 9 and other impurities and then washed with
water to obtain free 5 in 42% yield.10 The obtained 5 could
be used without further purification.
Although a few kilograms of 5 were synthesized by this
procedure for nonclinical use, the issue remains that this
procedure needs dichloromethane, which can be harmful to the
environment, as a reaction solvent. We therefore searched for
another procedure for the Friedel–Crafts reaction.
At first, we tried it in the absence of a solvent. Friedel–Crafts
reaction of 7 with benzylchloride 8 without solvent also gave
ketone 5 but posed two problems. (1) The reaction mixture
easily solidified and was not suitable for large-scale synthesis.
(2) On a 20-L scale, the three materials 7, 8 and ZnCl2 were
stirred and heated to start the reaction, but we were not able to
control it. A reaction calorimeter study of this reaction revealed
a large heat evolution of 327 kJ/mol and an adiabatic temper-
ature rise of 355 °C. In order to control the reaction, 7 was
added gradually into a mixture of 8 and ZnCl2, but this
procedure gave 5 in poor yield.
Conclusion
We have developed a convenient and economical one-step
synthesis of 5 starting from the easily available compounds 7
and 8. This procedure includes two different conditions. For
small-scale synthesis in a laboratory, the reaction by anhydrous
ZnCl2 in dichloromethane can be recommended. On the other
hand, for large-scale synthesis, the reaction with water should
be adopted to protect the environment and reduce the safety
risk. Modification of our procedures should be useful for
synthesizing 5-substituted-2-furyl ketones.
Experimental Section
General. Melting point was determined on Büchi apparatus
1
and is uncorrected. The H NMR spectra was measured on a
Varian Gemini 300 MHz FT NMR spectrometer.
5-(4-Fluorobenzyl)-2-furyl Methyl Ketone (5). Method A.
To a solution of 7 (19.71 g, 0.18 mol) in dichloromethane
(120 mL) were added 8 (42.9 mL, 0.36 mol) and ZnCl2
(36.6 g, 0.27 mol), and the mixture was allowed to reflux
with stirring. After 12 h the precipitated complexes with
5 and zinc chloride were collected and washed with
dichloromethane. Water was added to the complexes, the
mixture was extracted with ethyl acetate, and the organic
layer was washed with water and aqueous NaHCO3 and
then dried over anhydrous Na2SO4. The solvent was
removed in vacuo. n-Hexane was added to the residue,
and the precipitates were collected by filtration and dried
Method B. We investigated other solvents to replace
dichloromethane and found H2O to be the best (Scheme 3).
The Friedel–Crafts reaction in the presence of H2O is a two-
(6) (a) Tanaka, S.; Tomokuni, H. J. Heterocycl. Chem. 1991, 28, 991.
(b) Skouta, M.; Lesimple, A.; Le Bigot, Y.; Delmas, M. Synth.
Commun. 1994, 24, 2571. (c) Heaney, H.; Papageorgiou, G. Tetra-
hedron 1996, 52, 3473. (d) Uneyama, K.; Momota, M.; Hayashida,
K.; Itoh, T. J. Org. Chem. 1990, 55, 5364. (e) Yamauchi, M.; Shirota,
M.; Watanabe, T. Heterocycles 1990, 31, 1699. (f) Baudry, M.;
Barberousse, V.; Descotes, G.; Faure, R.; Pires, J.; Praly, J.-P.
Tetrahedron 1998, 54, 7431.
1
(7) (a) Fernandez, P. A.; Bellamy, T.; Kling, M.; Madge, D. J.; Selwood,
D. L. Heterocycles 2001, 55, 1813. (b) Koyanagi, J.; Yamamoto, K.;
Nakayama, K.; Tanaka, A. J. Heterocycl. Chem. 1995, 32, 1289. (c)
Moldenhauer, O. Liebigs Ann. Chem 1953, 580, 176. (d) Belen’kii,
L. I.; Gromova, G. P.; Kolotaev, A. V.; Krayashkin, M. M. Chem.
Heterocycl. Compd. 2000, 36, 256. (e) Barton, D. H. R.; Brown, B. D.;
Ridley, D. D.; Widdowson, D. A. J. Chem. Soc., Perkin Trans. 1 1975,
2069.
to give 5 (16.4 g, 42%). Mp 28 °C. H NMR δ (CDCl3):
2.43 (s, 3H), 4.01 (s, 2H), 6.09 (d, J ) 3.5 Hz, 1H),
6.96–7.26 (m, 5H).
5-(4-Fluorobenzyl)-2-furyl Methyl Ketone (5). Method B.
A mixture of 7 (187.2 g, 1.7 mol), 8 (491.6 g, 3.4 mol), aqueous
50% ZnCl2 (486.6 g, 1.785 mol) and water (131 mL) was
vigorously stirred at 85 °C. After 6 h ethyl acetate (1498 mL)
(8) (a) For reviews, see: Katritzky, A. R.; Taylor, T. AdV. Heterocycl.
Chem. 1990, 47, 102. (b) Sargent, M. V.; Dean, F. M. ComprehensiVe
Heterocyclic Chemistry; Pergamon Press: Oxford, 1984; Vol. 4, p 599.
(c) Heany, H.; Ahn, J. S. ComprehensiVe Heterocyclic Chemistry;
Pergamon Press: Oxford, 1996; Vol. 2, p 297.
(10) The reaction mixture was quenched with H2O and extracted with
dichloromethane or ethyl acetate extract. The yield was estimated by
HPLC analysis of the extract.
(9) Valenta, M.; Koubek, I. Collect. Czech. Chem. Commun. 1976, 41, 78.
1060
•
Vol. 11, No. 6, 2007 / Organic Process Research & Development

was added to the reaction mixture, and the organic layer was
washed with 3.65% aqueous HCl (936 mL × 3), 5% aqueous
NaHCO3 (936 mL) and water (936 mL). Each aqueous layer
was back-extracted with ethyl acetate (374 mL). The combined
organic layer was condensed in vacuo and then distilled under
reduced pressure at 115–144 °C (370–210 Pa). A mixture of
the distillate and isopropyl alcohol (144.4 mL) was allowed to
cool at 0 °C with stirring, and then seed crystals were added.
After 0.5 h n-hexane (288.7 mL) was added, and then the
mixture was allowed to cool at –10 °C with stirring. After 2 h
the obtained crystals were filtered, washed with chilled n-
hexane, and dried to give 5 (161.2 g, 43%).
Acknowledgment
We thank Dr. T. Konoike, Mr. Y. Kokuryo, and Mr. Y.
Sugata for their helpful discussion.
Received for review May 29, 2007.
OP700117Q
Vol. 11, No. 6, 2007 / Organic Process Research & Development
•
1061